a Parthian word of unknown derivation for “poet-musician, minstrel.”


GŌSĀN, a Parthian word of unknown derivation for “poet-musician, minstrel.” It was evidently much used, since it was borrowed into several neighboring tongues; but because the Parthian language almost disappeared and was not recovered to any extent until the 20th century, the term first came to notice through its use in an 11th-century classical Persian text, Faḵr-al-Din Asʿad Gorgāni’s Vis o Rāmin. There it occurs twice in one passage (ed. Todua and Gwakharia, pp. 300, l. 12 to 301, l. 24, episode 69 ll. 17-24 in the editions of Minovi and Maḥjub, tr. pp. 202-3). Read from the Arabic script as having initial “k,” it was first taken to be a proper name; but in the late 19th century Kerope Petrovich Patkanov (cited by von Stackelberg, p. 495) identified it from the context as a common noun meaning perhaps “musician,” and he suggested that it was an obsolete Persian term, traceable as a loanword in the well-attested Armenian gōsān. He read it nevertheless as “kusān”; but R. von Stackelberg proposed instead “gusan” and pointed out that Georgian mgōsāni was probably also derived from it.

The first indication that it was a Parthian word came when Harold Walter Bailey drew attention (1934, p. 11) to its occurrence in the Mojmal al-tawāriḵ (ed. Bahār, p. 69, tr. Mohl, pp. 514-15). There it is told that the Sasanian king Bahrām V Gōr asked the king of the Indians to send him gōsāns to make free minstrelsy (rāmeši) for his subjects when they drank wine, and the author explains that “in the Pahlavi language gōsān means ḵonyāgar.” Since the latter is a Persian word for “minstrel” (Mid. Pers. huniyāgar), it appears that “Pahlavi” is used here in its proper sense of “Parthian.” The text goes on to say that the “singers” who duly came from India, men and women, were the ancestors of the gypsies (luri) of the author’s own day, and that Bahrām gave them goods and animals “that they might without charge make minstrelsy for the poor.” Here an accurate definition of the calling of gōsān, portrayed in one of its aspects, has been confused by the identification of gōsāns with gypsies as popular entertainers. So far no occurrence of gōsān used for a woman singer is known, whereas such singers are found among gypsies.

Bailey, seeking an etymology for the word, was inclined to read it still with initial “k” (1934, p. 515; idem, 1954, p. 9, n. 6), but Walter Bruno Henning put an end to uncertainty on this point when he discovered (by putting together fragments of a manuscript page) a Manichean Parthian passage in which the word occurs written gwsʾn in the clear Manichean script (apud Boyce, 1957, p. 11). The text is attributable to the 4th or 5th century C.E, and runs: “like a gōsān, who proclaims the worthiness of kings and heroes of old, and himself achieves nothing at all” (čawāgōn gōsān kē hasēnagān šahrdārān ud kawān hunar wifrāsēd ud xwad ēwiž nē karēd). Henning (ref. apud Boyce, 1957, p. 17) also identified another borrowing of the Parthian word in Mandaean, where two occurrences of it, as gwsānā, had been brought together by Mark Lidzbarski (pp. 164, 166-67). One passage indicates that the gōsānā was a person of considerable social standing, on a level with a district judge. The other expresses the same scorn found in the Manichean Parthian passage for the minstrel as a man of words not deeds, and also, like the one in the Mojmal al-tawāriḵ (p. 69), assigns him a lowly status. The speaker has been entreated, with the offer of rich gifts, to sing a particular song, but refuses, declaring: “I am no gōsānā who makes music before humble people. I am a man whose words and songs are clubs and blows.” Susequently one or two occurrences of gōsān were noted in classical Persian texts, in passages where gōsāns are named disapprovingly among other low-class entertainers. The earliest of such occurrences is in the work of the 10th-century author Ḥamza Eṣfahāni (pp. 54-55, 63; the printed Arabic text has the word as ḥusiān, an apparent scribal error for jōsān, pl. jawāsena). Ḥamza also associated gōsāns with gypsyies. The word had probably gained some currency in the koine of late Middle Persian, and survived in limited general use for some time thereafter.

Meantime Vladimir Minorsky had established, in a series of studies, that Gorgāni’s Vis o Rāmin is derived, through a written Middle Persian version (made in the early Islamic period), from a Parthian oral work. In it, therefore, the occurrence of gōsān can be seen as a survival from the Parthian text, which Gorgāni could retain in this particularly significant passage because the meaning of the word would still be intelligible to his 11th-century readers.

It seems unlikely that any Parthian minstrel poems were set down in writing in early Sasanian times, and all that were recorded thereafter had necessarily to be rendered into Middle Persian, for it can be deduced (Boyce, 1979, p. 116) that at some point in the 4th century C.E. the Sasanians decreed that of the Iranian languages current within their empire only their own Middle Persian might be written, with the use everywhere of spoken Middle Persian being probably strongly encouraged. The training of Parthian scribes to write in their own distinctive script must therefore have ceased, and by the time of the Arab conquest in the 7th century knowledge of this script is likely to have been lost, with a block having been put on the development of Parthian as a written language. So when eventually the Parthian gōsān texts to be considered here were recorded, it could only be in Middle Persian; and it was only with the 20th-century recovery of the Parthian language that it became possible to identify in some of them a few distinctive Parthian words, idioms, and grammatical constructions; and this led on to the discovery in others of Parthian proper names, geographical settings, and political and social institutions characteristic of Arsacid times.

The result has been only a few texts, five in all, being identified as of gōsān origin; but because of the huge destruction of Zoroastrian manuscripts in Islamic times, these five comprise the greater part of what survives from what had been evidently a rich and abundant oral literature of entertainment. Of them two, the Ayādgār i Zarērān and the Draxt i āsurig (qq.v.) survive in the Middle Persian language and script, each having sufficient religious content to be copied repeatedly by Zoroastrian priests. The other three, with what religious content they had had reduced to a minimum, exist in classical Persian renderings. These are the pre-Goshtaspian parts of the Kayanian cycle, as celebrated by Ferdowsi in his Šāh-nāma; the romantic epic, Vis o Rāmin, in the version by Gorgāni; and the story of Bēžan o Manēža (see bèûan), which was probably by origin an independent romantic lay, known to Ferdowsi from an old Pahlavi (i.e., Mid. Pers.) book (daftar-e Pahlavi; Šāh-nāma, ed. Khaleghi-Motlagh, III, p. 306, l. 1), and incorporated by him in the Šāh-nāma. Few though they are, these works between them shed a fair amount of light on gōsāns and their craft.

Poetry and music, it is evident, were among the Parthians’ greatest pleasures, and for them the two were indissolubly joined (no native word is known in any Iranian language for a poet as distinct from a singer. Moreover, to compose and sing verses appears to have been a common polite accomplishment in Parthian society. So Rāmin, a prince and mighty warrior, constantly sings to his lute love-songs of his own making for Vis, or pours out poetic lamentations to vie with the nightingales; and young Bastwar (see BASTŪR), finding his father Zarēr dead on the battlefield, expresses his grief by extemporizing a poignant lament for him. (Phalavi Texts, ed. Jamasp-Asana, p. 12, ll. 1-3, para. p. 84) With minstrelsy thus cultivated by aristocratic amateurs, the standards demanded of gōsāns in the employ of kings and nobles must have been very high for them to please their listeners. But, as with musicians and poets in later times, wide variations in abilities must have existed among gōsāns generally, with diversity too in their audiences, which are reflected in the differences in status accorded them in the Mandaean passages cited above. The craft of the poet-musician was still practised at the highest levels throughout the Sasanian period, from which time the name of one or two individual minstrels are known, notably that of the great Bārbad (q.v.); and it was carried on to some extent into early Islamic times, in that there were poets then, among them the illustrious Rudaki and Farroḵi Sistāni, who were also accomplished musicians.

Nothing is known of how a gōsān was trained, but the predominance of the hereditary principle in ancient Iran makes it probable that the calling passed regularly from father to son, with training beginning at an early age; and this training must have been exacting, even regarding the craft of poetry alone, since this required much committing to memory of verses of all types—narrative, lyrical, laudatory, elegiac, satirical, gnomic—so that by this means apprentices acquired a store of traditional vocabulary, imagery, and themes by drawing on which they could create new compositions of their own; and they would also thus have learnt a quantity of familiar and well-loved poetry with which to please listeners on demand. Probably every apprentice, according to his talents and abilities, sought to build up as varied a repertoire as possible, to increase his earning power; and no doubt a man of exceptional talents could find service with a patron of his own choosing. Probably, however, each noble family would have had its own hereditary gōsāns, who knew its traditions and could celebrate in song its illustrious forbears and their achievements, thus sustaining family pride.

What emerges very strikingly from the passage cited above from Vis o Rāmin (ed. Todua and Gwakharia, pp. 300-2) is the amount of privileges a highly talented gōsān could expect to enjoy. King Mōbad is shown presiding over a festive gathering in a garden in spring, and the gōsān, standing near, is enhancing the company’s pleasure in its wine by “stringing the modes of those at a drinking bout (rāhhā-ye ḵošgovārān; ed. Todua and Gwakharia, p. 300, l. 13, tr. p. 203); but as the evening advances and all are a little flushed by drink, he dares to sing some charming verses which contain a dangerously provocative allegory, for they allude to the adultery of the king’s wife, Vis, with his brother Rāmin. Even in his cups Mōbad divines their meaning, but his rage flares up not against the gōsān but against Rāmin, on whom he springs, dagger in hand, to kill him.

Later in Vis o Rāmin minstrels (here called rāmeš-garān) appear at Rāmin’s wedding to the Lady Gol, when “great men gave largess freely and minstrels were much sought” (ed. Todua and Gakhwaria, p. 334, ed. Minovi and Maḥjub, episode 75, tr. p. 225). All sang in honor of Rāmin and praising Gol, and Gorgāni offers some of their verses, rich in word-play on the bride’s name, which means “Rose,” and in imagery of the rose (ed. Todua and Gakhwaria, p. 335, tr. p. 226).

All such occasional verses were forgotten, like most of the rest of gōsān poetry, with the development of a written literature and new conventions in versification; and what little survives consists of longer narrative poems. Of these by far the most important for their contribution to the literature and national consciousness of Muslim Persia, are the heroic lays recast by Ferdowsi. Reciting such lays defines the gōsān’s craft in the passages cited above the Parthian and one of the Mandaeans, and it was probably always a highly valued part of his activity. The times of the Arsacid Empire, itself part of a politically complex world and administered through a well-trained bureaucracy, cannot properly be characterized as a “heroic age”; but the conventions of such an age still shaped this genre of its literature. Valor and prowess in war were greatly prized, and their worthy celebration in verse, with stirring music, evidently gave keen pleasure; and those lays that concerned past deeds and events and had a basis in historical fact, or in legends and myths that were regarded as facts, would have had added an element of high seriousness. In lays concerning the Kayanians, there was also a religious element, with these Avestan figures being seen as the forbears of Zoroaster’s patron, Wištāsp (see GOŠTĀSP), and as accompanied by the royal glory (see FARR(AH)), which the Parthians believed was then attending their own Arsacid kings. From remote times, there had plainly been cross-fertilization between priestly lore and minstrel traditions; and in this instance it may well have been priests who, in the interest of exalting Kavi Wištāsp, created a genealogy for him from earlier kavis, who are always presented accordingly in a fixed succession in the Avestan yašts. This linear ordering was probably an essential factor for the formation of an Iranian epic. There are countries with a fine heroic literature which consists of individual lays that are never set together to form an epic narrative; but the putative action by Zoroastrian priests provided the basis for a continuous “history” of the Kavis (i.e., Kayanians), which was, it seems, adopted by minstrels and presumably celebrated by them over generations. But comparisons of later versions of Kayanian tales with brief allusions in the yašts suggest that much had been lost by the time these ancient traditions reached the Parthian gōsāns, and that the materials had worn a little thin, though enough remained to stir the imagination and gratify both religious and patriotic feelings. But the patriotic element was hugely increased when versions of the Kayanian cycle were composed, in which these ancient chieftains were surrounded by a trusted and heroic band of Pahlavāns, that is, Parthians. In time this term, its original meaning forgotten, came to be understood as “heroic, noble, splendid”; but by later Arsacid times a Parthian knew himself as a Pahlav or Pahlavān (cf. the identification of his language as pahlavānig izvān in Mir. Man. II, p. 303, R II 1), and to hear now of his fellow countrymen distinguishing themselves in the service of the divinely favored Kayanians must have evoked in him immense pleasure and pride. This development, it has been suggested, was initiated by a gōsān in the service of the Godarzian family (see GŌDAR) who found links between an Avestan figure of the Kayanian cycle and a famous ancestor, Gōdarz son of Gašwād, of his own patron (see Boyce, 2000, pp. 155-57; idem, in Boyce and Grenet, Zoroastrianism IV [in preparation], Ch. 9). Other such connections were then made out, and what had probably till then been separate lays celebrating the deeds of individual Godarzians in the distant past became blended with the linked ones telling of the still more remote Kayanians, and so what may be termed a Godarzian-Kayanian epic evolved. This survives in that part of Ferdowsi’s Šāh-nāma, which is generally considered to contain some of his finest poetry, inspired perhaps by the quality of the Parthian original that had come down to him through various stages.

The great sacred fire of Ādur Gušnasp (q.v.) in Azerbaijan, which was under Arsacid suzerainty, is mentioned a number of times in this part of the Šāh-nāma, figuring prominently in connection with the death of Afrāsiāb (q.v.); and one way by which the Godarzian-Kayanian epic began to gain national currency may have been through such topical verses being sung by gōsāns to pilgrim gatherings at its temple. It would seem that in the course of time gōsāns of other noble families also, notably that of Kārin, introduced their patrons’ forbears into various Kayanian tales; and at some point Rostam, the Sagzi (Saka), lord of Zābol, entered the thus evolving epic, to become its dominant heroic figure. That his introduction followed that of Gōdarz and his kin is recognized (Christensen, p. 139; Yarshater, p. 459), and a number of scholars have sought to connect it in some way with the great Parthian house of Surēn, which had hereditary estates in Zarang (OP Zranka; Nöldeke, Geschichte der Perser, p. 437, n. 5; see DRANGIANA). The Parthians of Zarang developed connections with the Sakas who settled in Sakastān/Sistān (the present-day province of Rōdbār in Afghanistan, see Markwart, pp. 35-36), which means that the Surēn family’s gōsāns would have had opportunities to hear songs about Rostam and to add these to their own repertoire; and these gōsāns can be assumed to have been highly talented, serving as they did such eminent masters. There was moreover in Zarang what was probably the most revered of all Zoroastrian holy places, the remarkable Kuh-e Ḵᵛāja, rising dramatically from the lake in which, it was believed, the future World Savior, the Saošyant, would be conceived (see ESCHATOLOGY i). Pilgrims from every Zoroastrian community are likely to have made their way there, and again it may have been at big yearly pilgrim gatherings that songs about Rostam were heard, to be memorized by visiting gōsāns and carried away by them to Zoroastrian audiences everywhere. It remains remarkable that Rostam, always in some measure an outsider, should have become in the national epic the Kayanians’ doughtiest champion and the greatest hero of Iran; and among diverse factors which must have contributed to this may well be that he entered Parthian tradition at this holiest of sanctuaries, and that this imbued him with a special importance. It seems, however, that those gōsāns who wove stories about him into the Iranian national epic were imitating the bold pioneering ventures of gōsāns of the Godarzian family, although without justification of perceived Avestan-Parthian links. His promotion probably went on well into Sasanian times, and he is even brought into the romantic tale of Bēžan o Manēža, although this damages its artistic integrity by a clumsy ending.

The introduction of an extraneous heroic element occurs also in Vis o Rāmin, with Vis, who belongs to Māh (Media), being given for father a “Qāran.” He almost certainly represents a head of the house of Kārin, which had hereditary estates in Media (Nöldeke, Geschichte der Perser, p. 437, n. 5), and as Qāran he dies heroically in battle, with fallen comrades all around him. This is the one incident in the remains of Parthian oral literature which can be plausibly identified with a known histori-cal event, namely the battle of Gotarzes II against the young pretender Meherdates, which took place in Māh/Media in 50 C.E., for in that battle Karin, the commander of Meherdates’ troops, died valiantly, fighting against odds (Tacitus, Annals 12.13-14; for the identification of Qāran with this Kārin, see Minorsky, 1947, p. 30; idem, 1964, p. 187). A lay would certainly have been composed at the time by a gōsān of the Kārin family to honor his brave stand, and this was presumably handed down by his descendants and eventually, to please the family, drawn on to embellish the locally popular tale of Vis and Rāmin. Otherwise two factors, ignorance of details of the wars fought by the Parthians on their northeastern frontier, and the way that the gōsāns plainly molded events to fit epic conventions, leave little hope that it will ever be possible to link deeds of the Pahlavāns with any certainty to actual historical happenings.



Harold Walter Bailey, “Iranica (II),” JRAS, 1934, pp. 505-18; repr. in idem, Opera Minora, ed. Mahyar Nawabi, 2 vols., Shiraz, 1981, I, pp. 11-24.

Idem, “Ariana,” in Erik Gren et al., eds., Donum Natalicium H. S. Nyberg Oblatum, Uppsala, 1954, pp. 1-16.

Mary Boyce, “The Parthian gōsān and Iranian Minstrel Tradition,” JRAS, 1957, pp. 10-45.

Idem, Zoroastrians: Their Religious Beliefs and Practices, London, 1979.

Idem, “Gotarzes Geopothros, Artabanus III, and the Kingdom of Hyrcania,” in Reinhard Dittmann et al., eds., Variatio Delectat: Iran und der Westen: Gedenkschrift Peter Calmeyer, Alter Orient und Altes Testament 272, 2000, pp. 155-66.

Arthur Emanuel Christensen, Les Kayanides, Copenhagen, 1931.

Mark Lidzbarski, ed. and tr., Das Johannesbuch der Mandäer, 2 vols. in 1, Giessen, 1915.

Faḵr-al-Din Asʿad Gorgāni, Vis o Rāmin, ed. Mojtabā Minovi, Tehran, 1314/1935; ed. Moḥammad-Jaʿfar Maḥjub, Tehran, 1337 Š./1958, repr. 1369 Š./1990; ed. Magali A. Todua and Alexander A. Gwakharia, Tehran, 1349 Š./1970; tr. George Morrison as Vis and Ramin, New York and London, 1972.

Vladimir Minorsky, “Vīs-u-Rāmīn, A Parthian Romance,” BSOAS 11/4, 1946, pp. 741-63; 12/1, 1947, pp. 20-35; repr. as “Vīs-u-Rāmīn (1946-7),” in idem, Iranica: Twenty Articles, Tehran, 1343 Š./1964, pp. 151-99, tr. Moṣṭafā Moqarrabi as “Vis o Rāmīn,” FIZ 4, 1335 Š./1936, pp. 3-73 (rev. and repr. in Maḥjub’s ed.).

Idem, “Vīs-u-Rāmīn,” BSOAS, 16/1, 1954, pp. 91-92; 25/2, 1962, pp. 275-86.

Mojmal al-tawāriḵ, partial tr.Jules Mohl, as “Extraits du Modjmel al-Tewarikh, relatifs à l’histoire de la Perse,” JA, 1841, pp. 515-16, 534.

R. von Stackelberg, “Lexicalisches aus ‘Wīs ō Rāmīn’,” ZDMG 48, 1894, pp. 490-97.

Ehsan Yarshater, “Iranian National History,” in Camb. Hist. Iran III, pp. 359-477.

(Mary Boyce)

Originally Published: December 15, 2002

Last Updated: February 17, 2012

This article is available in print.
Vol. Xi, Fasc. 2, pp. 167-170